Head-Fi.org › Forums › Equipment Forums › Sound Science › The Purpose of a Great Amplifier?
New Posts  All Forums:Forum Nav:

The Purpose of a Great Amplifier? - Page 2

post #16 of 26
Quote:
Originally Posted by bigshot View Post

What amp has a problem with channel separation? Every one I've seen has bleedthough well below the basement. And as for impedence and power vs THD, I've never seen any of that to be a problem either. As long as an amp is powerful enough to do the job, it's fine.


Yeah.

My point was to highlight the characteristics of the amp itself, whichever headphone is paired with it, the amp's output should have the same specs.

The combination of the amp + headphone is a different system.

 

Headphone amps *usually* have plenty of current. Current limiting is a greater issue with speakers IMO, because they go down to single digits of impedance.

post #17 of 26
Thread Starter 
Quote:
Originally Posted by proton007 View Post

I think you got the basics bang on. An amplifier is primarily an electrical device more than the artsy sound enhancer. And any measure of performance that applies to an electrical device applies here.

This is common knowledge, there are 4 aspects of an amplifier: frequency response, gain, noise, and distortion.

 

So an ideal amp would be: the same gain for all frequencies, with minimal noise and distortion.

 

These features can help find out the more neutral and cleaner amps out there. Generally, the better the specs, the better the amp is, but whether these differences are audible is another matter.

 

Now, the audible properties of these specs:

Frequency response : same as headphones (bass, treble, mids etc).

Gain: Amplification (volume increase).

Noise : Most commonly heard as hissing sound.

Distortion : Can be Harmonic Distortion, or Clipping.

 

Quote:
Originally Posted by Strangelove424 View Post

Measurements such as noise, total harmonic distortion + noise (THD+N), and to a lesser extent crosstalk, should be close to nothing... either hundreds or thousandths of a percent or, in the case of dB ratings, below the noise floor and out of audible range. Clean/neutral SS amps are easy to come by, but specs can often be incomplete, even on expensive audiophile gear. Professional equipment is less stricken by this affliction, but that's a different discussion for a different thread. If you want the most neutral amp you can get, try to find as many hard numbers as humanly possible, and balance that out with subjective impressions of the amp. 

Thanks. Below are the specs for the Schiit Magni, an amplifier I'm using for my HE-500s and which, to my ears, has no particular issues driving the 'phones. Based on what you have said and what I think I know about various specifications, this should be a pretty decently neutral/clean amp, right? 

 

I understand the basics of what high and low impedance means to headphones and higher and lower total wattage and how that happens to mean a high impedance set of headphones might need high voltage, but lower overall current. At the same time, orthos often are lower impedance, but need higher overall current to get the air moving. I've seen people make statements like "HE-500s need at least 1w of power". Is there any headphone specifications that might indicate this to be the case or is this kind of statement based on personal experience? When I look at specs for the HE-500s, for example, I see that they're 38ohm and have an 89DB efficiency. I don't see how I could make an overall current need statement based on those two numbers...

 

Frequency Response: 20Hz-20Khz, -0.1db, 2Hz-200KHz, -3dB


Maximum Power, 32 ohms: 1.2W


Maximum Power, 50 ohms: 1.0W


Maximum Power, 300 ohms: 260mW


Maximum Power, 600 ohms: 130mW


THD: Less than 0.005%, 20Hz-20KHz, at 1V RMS


IMD: Less than 0.007%, CCIF


SNR: Greater than 100db, unweighted, referenced to 1V RMS


Crosstalk: -70dB, 20 Hz-20KHz


Output Impedance: Less than 0.1 ohms


Gain: 5 (14db)

post #18 of 26

I believe there is a formula that you can use to figure out what power is needed to drive a headphone of a given sensitivity and impedence to a certain volume. I don't know it myself so hopefully someone here can share it.

 

If someone is saying a certain headphone needs a certain amount of power to be driven properly, they are likely going way overboard. It's really easy to tell if a planar magnetic headphone like the HE-500 is being driven with enough power. Got enough volume? Then you're good to go. Output impedence of your amp won't effect the frequency response at all for planars so volume is all you need to worry about in terms of power being delivered.

 

The power an amplifier is delivering depends on what the headphone is doing, it's not a consistant amount(note that specs are of maximum power output). If you have your volume knob only up half the way, you're not going to even come close to using all that power. I think some headroom is needed for complex passages, but there's absolutely no need to go so overboard that you're only using a 1/4 of your volume knob at most.

post #19 of 26

Headphone amps really aren't as necessary as people make them out to be. And there's no reason for a headphone amp to cost all that much. It's pretty simple stuff.

post #20 of 26
Quote:
Originally Posted by moses667 View Post

 

Thanks. Below are the specs for the Schiit Magni, an amplifier I'm using for my HE-500s and which, to my ears, has no particular issues driving the 'phones. Based on what you have said and what I think I know about various specifications, this should be a pretty decently neutral/clean amp, right? 

 

I understand the basics of what high and low impedance means to headphones and higher and lower total wattage and how that happens to mean a high impedance set of headphones might need high voltage, but lower overall current. At the same time, orthos often are lower impedance, but need higher overall current to get the air moving. I've seen people make statements like "HE-500s need at least 1w of power". Is there any headphone specifications that might indicate this to be the case or is this kind of statement based on personal experience? When I look at specs for the HE-500s, for example, I see that they're 38ohm and have an 89DB efficiency. I don't see how I could make an overall current need statement based on those two numbers...

 

Frequency Response: 20Hz-20Khz, -0.1db, 2Hz-200KHz, -3dB


Maximum Power, 32 ohms: 1.2W


Maximum Power, 50 ohms: 1.0W


Maximum Power, 300 ohms: 260mW


Maximum Power, 600 ohms: 130mW


THD: Less than 0.005%, 20Hz-20KHz, at 1V RMS


IMD: Less than 0.007%, CCIF


SNR: Greater than 100db, unweighted, referenced to 1V RMS


Crosstalk: -70dB, 20 Hz-20KHz


Output Impedance: Less than 0.1 ohms


Gain: 5 (14db)

 

 

The specs look very good. As far as the impedance thing is concerned, I feel that the importance of matching impedance/sensitivity of the speaker to impedance/output power of the amp is a remnant of large speaker systems, and has become less vital with solid state transistors and modern circuits that overload and fry themselves much less frequently. Those calculations are even less important for headphones, which are far easier to drive than speakers. The quick and dirty version: Amps are most efficient at 50% power output. So, can you get to acceptable listening levels at about 50% on the knob? Still got some headroom? Does it play loud enough? It can probably be made more complicated than that, but neither me nor my ears have the patience for all the math (and I have a sneaking suspicion that the vast majority of listeners don't either).     


Edited by Strangelove424 - 1/8/13 at 6:24pm
post #21 of 26
Quote:
Originally Posted by chewy4 View Post

I believe there is a formula that you can use to figure out what power is needed to drive a headphone of a given sensitivity and impedence to a certain volume. I don't know it myself so hopefully someone here can share it.

 

 

Yes, there is. It's described pretty well in the Wikipedia article http://en.wikipedia.org/wiki/Decibel.

 

So if you have a headphone that has a sensitivity of 89 db / 1 mW, at 1 watt you will have a dB level 89+30 or 119 db. That is above what you want to expose your ears to, and above any sound level an audience at a live performance should be exposed to.

 

A modern solid state amplifier that is operating within its design parameters should be a commodity product. Flat frequency response, inaudible noise and distortion and plenty of channel separation. Indistinguishable within its cohort in sound quality.

 

The tricky thing about headphones is that they have a large range of impedance and sensitivity values, so if you don't match the amp to the headphone you may be operating the amp outside its design parameters. For example an amp designed for a hard to drive phone may have too high a noise floor for IEMs.

 

Trying to pick an amp based on how it colors sound is madness. Use a device designed to tailor frequency response and add distortion instead. Sound engineers use this sort of thing all the time.

 

http://www.soundonsound.com/sos/apr10/articles/distortion.htm

 

Pick your amp based on features like connection options, sensitivity, impedance, power output, physical appearance, price, color, bling factor (for impressing dates/neighbors/members or your entourage), whether it will fit in your rack, warranty and the support history of the manufacturer. 


Edited by ehlarson - 1/8/13 at 7:32pm
post #22 of 26
Quote:
Originally Posted by moses667 View Post

 

Thanks. Below are the specs for the Schiit Magni, an amplifier I'm using for my HE-500s and which, to my ears, has no particular issues driving the 'phones. Based on what you have said and what I think I know about various specifications, this should be a pretty decently neutral/clean amp, right? 

 

I understand the basics of what high and low impedance means to headphones and higher and lower total wattage and how that happens to mean a high impedance set of headphones might need high voltage, but lower overall current. At the same time, orthos often are lower impedance, but need higher overall current to get the air moving. I've seen people make statements like "HE-500s need at least 1w of power". Is there any headphone specifications that might indicate this to be the case or is this kind of statement based on personal experience? When I look at specs for the HE-500s, for example, I see that they're 38ohm and have an 89DB efficiency. I don't see how I could make an overall current need statement based on those two numbers...

 

Frequency Response: 20Hz-20Khz, -0.1db, 2Hz-200KHz, -3dB


Maximum Power, 32 ohms: 1.2W


Maximum Power, 50 ohms: 1.0W


Maximum Power, 300 ohms: 260mW


Maximum Power, 600 ohms: 130mW


THD: Less than 0.005%, 20Hz-20KHz, at 1V RMS


IMD: Less than 0.007%, CCIF


SNR: Greater than 100db, unweighted, referenced to 1V RMS


Crosstalk: -70dB, 20 Hz-20KHz


Output Impedance: Less than 0.1 ohms


Gain: 5 (14db)


Take a look at this thread: http://www.head-fi.org/t/607282/headphone-amp-impedance-matching-basics-you-need-to-know  

 

There's a formula to calculate the power needed based on headphone sensitivity in dB/mW , Power = Antilog ( (Required SPL - SPL in dB/mW)/10)

If dB/V the following applies: Power = Antilog ( (Requiered SPL - SPL dB/V)/20)

 

At the required headphone impedance, if your amp should be able to provide the power.

post #23 of 26

In my setup my amps are ATTENUATORS more than amplifiers.  IE they do more signal attenuation than amplification.  Thats something I am always battling when I acquire a new amp or source, I am always trying to find ways to reduce output of the upstream component and reduce circuit gain of the downstream component.  My PPA and Millet for example all required resistor mods to get the circuit gain down to an appreciable level for my generally lower impedance cans.

 

The Millet colors the sound.  I run lower bias voltages (probably lower than most) to increase THD and flavor the sound.  My PPA is setup with OPA627s and discrete current buffers, its generally a "drier" (and more distant) sounding amp than the Millet.  The tube distortion from the Millet is definitely a more "wooly" sound, fatter and thicker in the middle.  I go back and forth between the two amps and my various cans from tune to tune as I run down the setlist.  There are pros and cons, but overall I don't think one is right and the other is wrong.


Edited by kramer5150 - 1/8/13 at 11:22pm
post #24 of 26

it does seem difficult to get across the idea that step down audio transformers may be more appropriate than amplification - nothing else can really preserve S/N with sensitive, low Z IEM - some reach 120 dB SPL with <200 mVrms - consumer desktop source is 2 Vrms

post #25 of 26
Quote:
Originally Posted by moses667 View Post

I frequently read about how different amplifiers change the sound of headphones in different ways... shouldn't a really great amplifier do NOTHING to the signal except increase the amplitude of the signal to make it louder? 

 

If there are changes to the signal, since this isn't a headphone producing noise to our ears, but electrical signal, shouldn't it be fairly easy to measure the input and output signal and see how closely they match? How close the amplifier comes to the ideal (assuming that is, in fact, the ideal)?

 

First, I entirely agree that amplifiers should "do NOTHING to the signal except increase the amplitude of the signal to make it louder". That is exactly what they should do.

 

In reality that is unachievable but it should be the aspiration.

 

Good amplifier manufacturers will pursue this objective, but knowing that there will be signal loss and distortion to a certain extent, they will tailor this to be as unobjectionable as possible.

 

Unfortunately throughout the 90s and the 00s we have seen a very significant increase in the number of amplifiers that "do stuff". By this I mean amplifiers that add "warmth" or other forms of distortion.

 

I suggest to people that they avoid these amplifiers and buy amplifiers that do the minimum to the input signal.

 

Amplifiers that are as transparent as possible will bring you all the excitement in the music, if it is present, they will bring any warmth in the music, if it is present, they will bring coldness in the music, if it is present. You get the idea I am sure.

 

I say, leave the music making to the musicians, that is their job. Hi Fi needs to get out of the picture, that is its job.


Edited by p a t r i c k - 1/9/13 at 1:01pm
post #26 of 26
Quote:
Originally Posted by p a t r i c k View Post

First, I entirely agree that amplifiers should "do NOTHING to the signal except increase the amplitude of the signal to make it louder". That is exactly what they should do.

In reality that is unachievable but it should be the aspiration.

Used to call that "A straight wire with gain". Technically unachievable, but in practice you can get close enough. Using an particular topology deliberately for its flaws always seemed like making "art" out of something that should be more of a technical function, but whatever floats the boat. I spent decades trying to reduce record surface noise to an inaudible level, then along comes the "low-fi" genre and deliberately mixes it back in. Yea, art. Whatever. I'll have mine straight up (with gain).
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Sound Science
Head-Fi.org › Forums › Equipment Forums › Sound Science › The Purpose of a Great Amplifier?